Our overall goal is to validate de novo DNA cytosine-C5 methyltransferases (MTases) as appropriate molecular targets for anti-cancer drug development and to propose and test new methods for screening specific inhibitory ligands of these enzymes. There is much evidence suggesting that alterations in DNA methylation are associated with carcinogenesis. Genome-wide decreases in 5mC content occur early in tumor development. However, specific regions of "hypermethylation" occur and are associated with epigenetic silencing of tumor suppressors or enzymes involved in DNA repair. This leads to the assumption that blocking or reversing hypermethylation will lead to reactivation of genes critical to inhibiting progression, restoring a normal phenotype or growth regulation or inducing apoptosis. Incorporation of 5-azacytidine or 5-aza-2'- deoxycytidine (5azadC, Decitibine) into DNA leads to DNA MTase inhibition with passive loss of methylation during cell division and activation of hypermethylated genes in cultured human cancer cells. Both drugs have been utilized in a variety of clinical trials as potential anti-cancer therapeutics. However, they are both cytotoxic and genotoxic. We have already developed and optimized non-genotoxic oligodeoxyribonucleotide (ODN) inhibitors of DNMT1, the most abundant form of DNA MTase in somatic cells. DNMT1 maintains methylation patterns established by de novo MTases during development. It has limited de novo activity in vitro and may be completely inactive in vivo. Thus, it is most probable that some member of the recently discovered DNMT3 family of DNA MTases, which can catalyze de novo methylation in vivo, is responsible for carrying out the establishment of methylation patterns and critical silencing of genes during tumorigenesis. The hypothesis we will test is that inhibition of de novo DNA MTases is necessary, not only for efficient reduction of methylation and activation of silenced genes but also for prevention of "de novo remethylation". Our specfic aims are: 1) To validate DNMT3a/3b or similar as yet unidentified de novo DNA MTases as targets by: a) Designing and optimizing small ODN inhibitors specific for DNA MTases 3a, 3b and inactive against DNMT1. They will contain 5-azaC, 5FdC, or abasic sites as replacements for the target C's in completely unmethylated recognition sites. b) Determining/comparing the effect of DNMT 3a, 3b inhibitors (alone and in combination with 5azadC or ODN inhibitors for DNMT1) on DNA methylation in cultured human cancer cells. This will be accomplished with a genome-wide CpG island screen and by bisulfite sequencing of CpG islands in genes known to be inactivated by methylation. c) Determining/comparing the effect of DNMT 3a, 3b inhibitors (alone and in combination with 5azadC or ODN inhibitors of DNMT1) on gene expression profiles using expression arrays for know tumor suppressors, cell cycle regulators and mediators of apoptosis. 2) Use a novel SDS-PAGE gel assay that measures inhibitor binding affinity for DNMTs to predict the potency of a variety of new inhibitory targets for DNMTs 1, 3a and 3b.